B.ACKERMAN, T. A. J O R D A N ,
4444 [CONTRIBUTION FROM
THE
c. R.EDDYA N D D. SWERN
VOl. 78
EASTERX REGIONAL RESEARCH LABORATORY~ A N D THE FATTY ACID DIVISIONOF OF AMERICAS SOAP AND GLYCERINE PRODUCERS, Inc.]
THE
ASSOCIATION
Phosphorus Derivatives of Fatty Acids. I. Preparation and Properties of Diethyl Acylphosphonates B Y BERNARD xCKERMAW,3
T.
x. J O R D A N , 3 c. ROLANDEDDYA X D DANIELS W E R N
RECEIVED J.4NUARP 18, 1956 Diethyl acylphosphonates have been prepared by the reaction of acyl chlorides with triethyl phosphite These mere dcrived from the even-numbered CI-CI8saturated fatty acids and oleic acid. Boiling point, density and refractive index were determined for each compound The diethyl acplphosphonates are unstable to moisture, the rate of hydrolysis decreasing with increase in size of acyl chain and phosphorus ester group
-1
The analogous reaction of trialkyl phosphites with alkyl halides (equation 1) and the structure proof of Kabachnikj suggest that form 11, the true phosphonate structure, is the more likely. I n the course of certain physical studies, we have found additional evidence that such is the case. The observed values of molar refraction (Table I) were found to agree much more closely with the values calculated on the basis of a phosphonate structure than with those based on an anhydride structure. The latter values can be obtained by the addition of 2.20 to the calculated acylphosphonate values appearing in Table I. The observed [,M.R.] values are higher than the calculated ones by a constant increment. Exaltation of the molar refraction is known to occur in compounds having conjugated systems; i t is likely that conjugation of the carbonyl and phosphonyl linkages accounts for the exaltation in this case. That the increment cannot be traced to the phosphonate group alone is shown by the observed [M.R.] values for the alkylphosphonates, which agree almost exactly with the calculated values. The diethyl acylphosphonates exhibited welldefined polarographic waves having half-wave potentials of -1.4 v. ZK. the saturated calomel electrode. On the other hand, diethyl dodecylphosphonate showed no polarographic reduction. It was assumed, therefore, that reduction of the acylphosphonates was due to the presence of the carbonyl group. Since the isolated carbonyl is not reducible in the applied voltage range of the electrolyte used (0 to -2.0 v.),I0it is likely that reduction is effected because of conjugation between the carbonyl and phosphonyl groups. The anhydride structure is not conjugated ; thus polarographic
reduction is additional evidence for the phosphonate form. Ultraviolet absorption spectra were obtained for diethyl octanoyl- and palmitoylphosphonates. Both samples showed a double peak a t 333 and 341 mk. The molar absorptivities were 59.04 and 62.81 for the CSand C16 derivatives, respectively. Infrared spectra were obtained on the diethyl acylphosphonates having 10, 12, 14 and 16 carbon atoms in the acyl groups. All of these were liquids and had spectra very similar to that shown in Fig. 2 . Differences among these four spectra were only those minor variations which would be expected from the differences in chain length. Tab!e I1 lists the absorption bands which may be associated with the phosphorus-containing part of the molecule, showing the degree of variation of band positions among these four compounds. The assignments are based on those of Be1lamy.l’ The 794 cm.-l band is also tentatively assigned to P-0-C
(10) C. 0. Willits, C. Ricciuti, H. B. Knight a n d D. Swern, A n a l . Chem., 24, 788 (1932).
(11) L. J , Bellamy, “ T h e Infrared Spectra of Complex hlolecules,” John U‘ilcy and Sons, Inc., New York, N. Y . , 1 9 X .
0 E
f 40 I
20
0
0
20
40 TIME,
60,
80
100
HOURS.
Fig. 1.-Rates of hydrolysis of dialkyl acyl- and alkylphosphonates in dioxane-water a t 93 =!c 1’.
4446
$
B. ACKERMAN, T. A.
JORDAK,
C.R.EDDY.IND D. SKERK
Yol.
is
the residual solvent was evaporated under vacuum. The residue was distilled under high vacuum yielding 52.3 g. (85%) of 2-bromoethyl laurate, b.p. 114-119' (0.5 m m . ) ~ Dodecyl chloride (Hooker Electro-Chemiand n 3 0 1.4564. 60 cnl Company) war, fractionally distilled, and the portion disci * tilling at 134-135' (11 mm.) was retained. 40 Diethyl Acy1phosphonates.-The preparation of diethyl 9 20 lauroylphosphonate is typical; t o lauroyl chloride (43.8 g., 0.20 mole) 39.8 g. (0.24 mole) of triethyl phosphite mas c = 0 added dropwise over a period of about one hour, while a 6 7 8 9 10 11 stream of dry nitrogen was passed through the solution for agitation and t o sweep out the ethyl chloride which formed. Wave length, p. The reaction was exothermic and temperature was controlled simply by the rate of addition of the triethyi phosphite. Fig. 2.-Infrared absorption spectrum for diethyl myrisT h e temperature was not allowed t o exceed 45 . After an toylphosphonate, in liquid state, 0.02 mm. thick, 98-1000/, additional hour, the reaction mixture was vacuum-distilled pure. Numbers beneath the bands indicate wave numbers, through a heated Claisen head, yielding 26.2 g. of diethyl in em.-'. I n addition t o the bands shown, the spectrum lauroylphosphonate, b.p. 151-155' (0.1 m m . ) and n 3 0 ~ 1.4388-90, as a colorless liquid. This product was later recontains strong, sharp bands a t 2854 and 2925 cm.-' and distilled with the main portion having an X30D of 1.4399broad, weak bands at 723 and 794 cm.-*. 1.4401. Diethyl acylphosphonates with molecular weights higher vibration in accordance with the suggestion of than t h a t of the lauroyl derivative were distilled in a n alemMeyrick and Thompson.I2 bic apparatus. I n every case a considerable quantity of unPresence of the strong band a t 1258 cm.-l is identified high-boiling residue remained in the distillation evidence that the compounds prepared have the flask. T h e physical properties of the diethyl acylphosphonates phosphonate structure, with a P-0 group, rather summarized in Table I. than the mixed anhydride structure, which would areDibutyl Lauroy1phosphonate.-Fifty grams (0.2 mole) of not contain a P+O group. The low frequency of tributyl phosphite was added dropwise t o 43.8 g . (0.2 mole) the carbonyl vibration a t 1698 cm.-' is also sug- of lauroyl chloride in about one hour. The reaction mixture gestive of conjugation between the C=O and was held a t 80-90" and dry nitrogen was passed through the mixture in order to remove the butyl chloride which formed P+O, providing further evidence that there is no as a by-product. After an additional hour the reaction mixoxygen atom between the acyl group and the ture was vacuum-distilled and 34.2 g. of crude dibutyl lauroylphosphonate, b.p. 168-182" (0.3 mm.) and n3'D 1.4430phosphorus atom. 35, was obtained. This was redistilled in a n alembic still, yielding about 20 g. of product of much higher purity, b.p. TABLE I1 150-155' (0.14 mm.) and n30g 1.4432. ASSIGNMEXTOF ABSORPTIONBAXDSOF DIETHYL ACYLAnal. Calcd. for C ~ O H ~ I O IP, P : 8.23. Found: P, 8.36. PHOSPHOSATES, c M . - ' Di-2-ethylhexyl Lauroylphosphonate .--Ninety-two grams 794 P-0-c (0.22 mole) of tri-2-ethylhexyl phosphite was added dropwise in one hour to 43.8 g. ( 0 2 mole) of lauroyl chloride 972-973 P-0 (pentavalent phosphorus) through which dry nitrogen was being passed. The reaction 1023-1025 P-0-C (aliphatic) was run a t room temperature, since raising the temperature 1164 P-0-Et hpl t o aid in the removal of 2-ethylhexyl chloride might have 1258 P-tO caused this by-product to react with the tri-2-ethylhexylphosphite. A4troom temperature the reactivity of the by1697-1698 c=o product was assumed to be negligible relative t o t h a t of the lauroyl chloride. After a n additional reaction time of one Experimental hour, the reaction mixture was distilled in a n alembic still. Trialkyl Phosphites.-Triethyl phosphite was obtained Only 11.2 g. of di-2-ethylhexyl lauroylphosphonate, h.p. from the Eastman Kodak Company and the Virginia165-188' (0.1 mm.) and n 3 0 ~ 1.4468, was obtained. The Carolina Corporation. Before use, it was distilled at atmoshigh boiling residue was present in much greater than usual pheric pressure, the portion distilling at 151-155' being amounts. retained. Tri-n-butyl phosphite was the Eastman Kodak Anal. Calcd. for C28Hj?O4P: P, 6.34. Found: P, 6.83. Company Yellow label grade. Tri-2-ethylhexyl phosphite, obtained from the Virginia-Carolina Corporation, was disAlkyl Phosphonates. Diethyl Dodecy1phosphonate.-Dotilled, the portion boiling at 148-150" (0.25 mm.) being decyl chloride (61.4 g., 0.3 mole) and triethyl phosphite retained. (52.4 g., 0.315 mole) were mixed in a flask fitted with a n air Acyl Chlorides.-The acyl chlorides, with the exception of condenser and the mixture was heated t o reflux (175'). decanoyl, oleoyl and stearoyl, were Eastman Kodak Co. Refluxing was continued intermittently for 27 hours, during organic chemicals. Decanoyl chloride was prepared by re- which time the temperature rose slowly to 225'. At this acting highly purified decanoic acid with an excess of oxalyl time, there mas no evidence of any material refluxing and the chloride. Oleoyl chloride was prepared similarly, using heating was discontinued. The mixture was vacuum disoleic acid prepared from olive oil fatty acids by low temperatilled and 53.9 g . (67y0) of diethyl dodecylphosphonate, b.p. ture crystallization and fractional di~ti1lation.I~Stearoyl 137-139' (0.3 m m . ) and n Z 51.4389-91, ~ was obtained. chloride was prepared by treating highly purified stearic A n d . Calcd. for C I ~ H ~ L OP, ~ P10.11. : Found: P, 9.77. acid (m.p. 68-69') with a n excess of thionyl chloride. I n each of these cases, the reaction mixture, after removal of grams Diethyl Lauroxyethy1phosphonate.-Fifty-two excess chlorinating agent, was used without further purifica(0.17 mole) of 2-bromoethpl laurate was mixed with 58 g. tion for the preparation of the corresponding acylphosphon(0.36 mole) of triethyl phosphite and the solution was reate. fluxed (155'). The n 3 0 ~ of the mixture increased from a n Alkyl Halides.-2-Bromoethyl laurate was prepared by original value of 1.4292 to 1.4316 after 4 hours of refluxing refluxing 40 g . (0.2 mole) of lauric acid, 75 g . (0.6 mole) of and then remained constant for a n additional 10 hours. %-bromoethanol and 2.3 g. of concentrated sulfuric acid in T h e reaction mixture was then vacuum distilled, yielding 100 ml. of toluene for 3.5 hours, with azeotropic removal of 35.2 g . (57aj0) of diethyl lauroxyethylphosphonate, b.p. water. -4fter cooling, the dark-colored reaction mixture 162-164" (0.3 mm.) and n3'D 1.4419. was washed with water, dried by azeotropic distillation and A n d . Calcd. for C U H ~ ~ O SP, P : 8.50. Found: P, 8.63. Hydrolysis Experiments.-One to t n o grams of the phos(12) C . I. hleyrick and H. W. Thompson, J. Chem. Soc., 225 (1950). (13) H. R. Knight, E . F. Jordan, Jr., E. T. Roe a n d D . Swern, phonate w t s weighed accurately into a 100-ml. volumetric Biocheinical P v e p a v a i i o i i s , 2, 100 (1952). fl'i-li. Ai hot wlution of water:dioxane (30:70) \\as added 100 0' 80
'E
ESTERIFICATION O F DIPHENYLPHOSPHINODITHIOIC ,ACID
Sept. 5, 1956
to the mark and the contents mixed th0rough1y.l~ T h e solution was transferred to a reaction flask and heated t o reflux (93 zk 1") as quickly as possible. Refluxing was continued for 3-4 days (7 days in the case of di-2-ethylhexyl lauroylphosphonate). Ten-milliliter aliquots were removed by pipet frequently during the first 8 hours, then once a day until hydrolysis was complete. The aliquot was transferred t o 60-70 ml. of absolute ethanol and titrated immediately with 0.1 N NaOH. Polarographic Studies.-Details of the polarographic procedure and the solvent system have been reported in previous p u b l i c a t i o n ~ . ~ ~Methanol-benzene *~~ was the only solvent system used. Infrared Studies.-Infrared absorption spectra were obtained with a Beckman IR-3 spectrophotometer, using so(14) Twenty ml. additional dioxane was necessary in t h e case of di-2-ethylhexyl lauroylphosphonate because of incomplete solubility in t h e above solution. (15) W. E. Parker, C. Ricciuti, C. L. Ogg and D. Swern, TBIS JOURNAL, 77, 4037 (1955).
4447
dium chloride prisms. For the pure liquid samples a spacer apprctximately 0.02 mm. thick was used between two sodium chloride windows. A single sodium chloride window was used for the tape-recorded blank. Ultraviolet Studies.-The instrument was the Carey Model 11 recording spectrophotometer. Samples were dissolved in isooctane a t levels of 1 and O.l%, and measured in a cell of 1 cm. thickness.
Acknowledgment.-The authors wish to thank Dr. Constantine Ricciuti for the polarographic studies, Mrs. Ruth B. Kelly, Mrs. Katherine &I. Zbinden and Dr. Clyde L. Ogg for the analytical work, and Miss Anne M. Smith and Dr. George C. Nutting for the ultraviolet spectroscopy. The authors also wish to thank Mi-. E. T. Roe for helpful discussions. PHILADELPHIA, PA.
[CONTRIBUTION FROM THE CHEMICAL RESEARCH LABORATORIES, THELUBRIZOL CORPORATION]
Aromatic Phosphinic Acids and Derivatives. 11. Direct Esterification of Diphenylphosphinodithioic Acid1 BY T. ROBERTHOPKINS AND PAUL W. VOGEL RECEIVED MAY7, 1956 Alkyl diphenylphosphinodithioates have been prepared in good yields b y the direct esterification of diphenylphosphinodithioic acid with certain alcohols in the absence of a catalyst. The reaction proceeds readily with primary alcohols, higher boiling secondary alcohols and tertiary alcohols. Phenol yields a monothioester.
Very little work has been reported on the chemistry of the diarylphosphinodithioic acids and their derivatives. The first member of this series, diphenylphosphinodithioic acid, can now be conveniently prepared by the reaction of benzene with phosphorus pentasulfide in the presence of anhydrous aluminum chloride.2 Esters of this acid have not been reported previously. I n the present investigation i t was found that diphenylphosphinodithioic acid can be esterified directly either by addition of the thiol group to an olefin or by reaction with certain alcohols in the absence of a catalyst. The latter is a novel reaction in that water, rather than hydrogen sulfide, is elimina.ted and a dithioester is formed. Esters possessing primary alkyl groups are obtained from primary alcohols. An 8570 yield of the n-octyl ester was obtained from the reaction of noctyl alcohol with the acid a t 180". The alcohol was not first dehydrated since, if this were the case, the 2-octyl ester would result. This was demonstrated by the reaction of diphenylphosphinodithioic acid with 1-octene to give a 90% yield of the 2-octyl ester as shown in reaction (2). The addition of the thiol group to the olefin followed Markownikoff's rule. The structures of the esters were proven by saponification to the corresponding mer-
+ T Z - C ~ H I ~+ O H Ph2PSS-CsHIy + H20 (1) + 1-octene +Ph2PSS-CH(CH3)-C6HI8 (2) NaOH PhzPSSR + HzO PhzPOSNa + RSNa (3)
PhzPSSH PhzPSSH
-
(1) Presented before t h e Division of Organic Chemistry a t t h e 125th A.C.S. Meeting, Kansas City, Mo., March, 1954. (2) W. Higgins, P. Vogel and W. Craig, THISJ O U R N A L , 77, 1864 (19%).
captan and diphenylphosphinothioic acid. The mercaptans were identified by their 2,4-dinitrophenyl thioethers prepared by the procedure of The esterification of carboxylic acids requires the removal of a hydroxyl group from the acid. The relative order of reactivity of alcohols in this type of esterification is RI > RII > RII1. When a strong acid is esterified, the hydroxyl group is believed to be removed from the alcohol. No definite proof of this is known except in the case of the hydrogen halide acids where there is no alternative. The relative reactivity rates for alcohols in the "esterification" of these acids are in the order R'I' > R'I > RI. Diphenylphosphinodithioicacid may be classified as a strong acid, and as such it removes the hydroxyl group from the alcohol during esterification. One would, therefore, expect the relative reactivity rates of the alcohols to be RII1> RI1 > R'. Actually, the order seems to be RIII 7 R1 > RI1. Primary alcohols react readily while the lower secondary alcohols give little or no reaction a t their boiling points. Only one tertiary alcohol, t-butyl alcohol, has been studied. The reactivity rates of primary, secondary and tertiary alcohols were compared by the esterification of a series of butyl alcohols. %-Butyl alcohol reacted with the acid a t 117" for 28 hr. to give a 63'% yield of the n-butyl ester. When sec-butyl alcohol was employed, little or no reaction occurred after 30 hr. a t the reflux temperature (loo"), and 90% of the starting acid was recovered unchanged. The fact that t-butyl alcohol gave a 66% yield of the ('3) R W. Bost, J (1932).
D Turner and R D Norton, rbid, 54, 1985
